Classifying and Analyzing Physical Activities through Heart Rate Variability and Other Physical Metrics Using Holter Monitor Data
Classifying and Analyzing Physical Activities through Heart Rate Variability and Other Physical Metrics Using Holter Monitor Data

Summary/Abstract: This case study focuses on labeled data derived from a Holter monitor, a device that continuously measures and records heart activity, along with several other physical metrics. The primary objectives are twofold: first, to identify and explain similarities among the different types of physical activities based on the given metrics; second, to develop a predictive model that can accurately classify the type of activity given the recorded data. The research involves a detailed data preparation, including cleaning and extracting features, visualizing data by applying t-SNE and UMAP methods, and employing ensemble approach to build a model showing high precision, recall, and F1-scores in classifying various activities. This comprehensive approach demonstrates the efficacy of machine learning techniques that can be applied in health informatics and human activity recognition.

• Details
• Contents